The present invention relates to an improved process for obtaining a precursor of insulin or an insulin derivative having correctly bonded cystine bridges in the presence of cysteine or cysteine hydrochloride and a chaotropic auxiliary.
Human insulin is a protein with two amino acid chains together having 51 amino acid residues. Six cysteine residues are found in the two amino acid chains, each two cysteine residues being bonded to one another via a disulfide bridge. In biologically active human insulin, the A and B chains are bonded to one another via two cystine bridges, and a further cystine bridge is found in the A chain. Within a human insulin molecule, looked at statistically, there are 15 possibilities for the formation of disulfide bridges. In biologically active human insulin, only one of the 15 possibilities is found. The following cysteine residues are linked to one another in human insulin:
A 6-A 11
A 7-B 7
A 20-B 19
The letters A and B represent the respective insulin amino acid chain and the numbers represent the position of the amino acid residue, which is counted from the amino to the carboxyl end of the respective amino acid chain. Disulfide bridges can also be formed between two human insulin molecules such that an incalculable number of different disulfide bridges can easily result.
A known process for the preparation of human insulin is based on the use of human proinsulin. Human proinsulin is a protein having a linear amino acid chain of 86 amino acid residues, the A and B chains of the human insulin being bonded to one another via a C peptide having 35 amino acid residues. The formation of the disulfide bridges found in human insulin takes place via an intermediate, the cysteine residues of the human insulin being provided with a sulfur protective group, e.g. an S-sulfonate (xe2x80x94Sxe2x80x94SO3xe2x80x94) group (EP 0 037 255). A process for obtaining proinsulin having correctly bonded cystine bridges is additionally known (Biochemistry, 60, (1968), pages 622 to 629), which starts from proinsulin obtained from porcine pancreas, in which the cysteine residues are present as thiol residues (xe2x80x94SH). The term xe2x80x9ccorrectly bonded cystine bridgesxe2x80x9d is understood as meaning the disulfide bridges which are found in biologically active insulin from mammals.
Recombinant DNA processes allow precursors of insulin or insulin derivatives, in particular human proinsulin or proinsulin which has an amino acid sequence and/or amino acid chain length differing from human insulin, to be prepared in microorganisms. The proinsulins prepared from genetically modified Escherichia coli cells do not have any correctly bonded cystine bridges. A process for obtaining human insulin using E. coli (EP 0 055 945) is based on the following process steps:
Fermentation of the microorganismsxe2x80x94cell disruptionxe2x80x94isolation of the fusion proteinxe2x80x94cyanogen halide cleavage of the fusion proteinxe2x80x94isolation of the cleavage product having the proinsulin sequencexe2x80x94protection of the cystine residues of proinsulin by S-sulfonate groupsxe2x80x94chromatographic purification of the S-sulfonatexe2x80x94formation of the correctly bonded cystine bridgesxe2x80x94desalting of the proinsulinxe2x80x94chromatographic purification of the proinsulin having correctly bonded cystine bridgesxe2x80x94concentration of the proinsulin solutionxe2x80x94chromatographic purification of the concentrated proinsulin solutionxe2x80x94enzymatic cleavage of the proinsulin to obtain human insulinxe2x80x94chromatographic purification of the resulting human insulin.
Disadvantages of this process are the number of process steps and the losses in the purification steps, which lead to a low yield of insulin. From the stage of the isolated fusion protein via cyanogen halide cleavage, sulfitolysis and purification of the proinsulin, a loss of up to 40% of the proinsulin product has to be expected (EP 0 055 945). Similarly, high losses can occur in the course of the subsequent required to obtain the final product.
Yield increases in the preparation of human insulin or insulin derivatives by recombinant DNA means can be achieved if the number of process steps necessary can be significantly reduced.
EP 0 600 372 A1 (or U.S. Pat. No. 5,473,049) and EP 0 668 292 A2 disclose an improved process for obtaining insulin or insulin derivatives, in which the insulin precursor or precursor of the insulin derivative whose cystine bridges are not present in correctly linked form is reacted in the presence of a mercaptan, for example cysteine, and of at least one chaotropic auxiliary, for example urea or guanidine hydrochloride, to give an insulin precursor or precursor of the insulin derivative having correctly bonded cystine bridges. In the known process, these proteins are first dissolved in a very low concentration in aqueous solutions of a chaotropic auxiliary or of mixtures of various chaotropic auxiliaries. The protein mixture is then mixed with an aqueous mercaptan solution.
Surprisingly, it has now been found that the yields of correctly folded precursors of insulins or insulin derivatives can be increased and the reaction times for the folding process can be reduced by delaying the dissolution of the precursor in a first step by means of the chaotropic auxiliary. Instead, a mercaptan, such as cysteine or cysteine hydrochloride, is introduced into the aqueous suspension of the precursor in the first step, followed by dissolution of the precursor in a subsequent step by introducing a chaotropic auxiliary, and finally bringing about the correct folding of the precursor by dilution of the mixture to a preferred cysteine or cysteine hydrochloride concentration with an appropriate amount of water.
Accordingly, the present invention relates to a process for obtaining a precursor of insulin or an insulin derivative having correctly bonded cystine bridges in the presence of cysteine or cysteine hydrochloride and of a chaotropic auxiliary, which comprises successively carrying out the following steps:
(a) mixing an aqueous suspension of the precursor of insulin or an insulin derivative with an amount of cysteine or cysteine hydrochloride which results in approximately 1 to 15 SH residues of the cysteine or cysteine hydrochloride per cysteine residue of the precursor,
(b) introducing the cysteine- or cysteine hydrochloride-containing suspension of the precursor into an approximately 4 to 9 molar solution of the chaotropic auxiliary at a pH of approximately 8 to 11.5 and a temperature of approximately 15xc2x0 C. to 55xc2x0 C., keeping the mixture obtained at this temperature for approximately 10 to 60 minutes; and
(c) introducing the mixture at a pH of approximately 8 to 11.5 and a temperature of approximately 5xc2x0 C. to 30xc2x0 C. into an amount of water which results in a dilution of the concentration of the cysteine or of the cysteine hydrochloride in the mixture to approximately 1 to 5 mM and of the chaotropic auxiliary to approximately 0.2 to 1.0 M.
Preferably, the process is one wherein in step (a) the amount of cysteine or cysteine hydrochloride corresponds to an amount which results in approximately 1 to 6 SH residues of the cysteine or cysteine hydrochloride per cysteine residue of the precursor,
in step (b) the cysteine- or cysteine hydrochloride-containing suspension of the precursor is introduced into an approximately 4 to 9 molar solution of the chaotropic auxiliary at a pH of approximately 8 to 11 and a temperature of approximately 30xc2x0 C. to 45xc2x0 C. and the mixture obtained is kept for approximately 20 to 40 minutes at this temperature; and
in step (c) the mixture is introduced at a pH of approximately 8 to 11 and at a temperature of approximately 15xc2x0 C. to 20xc2x0 C. into an amount of water which results in a dilution of the concentration of the cysteine or of the cysteine hydrochloride in the mixture to approximately 1 to 5 mM and a concentration of the chaotropic auxiliary of approximately 0.2 to 1.0 M.
Chaotropic auxiliaries are compounds which break hydrogen bridges in aqueous solution, for example ammonium sulfate, guanidine hydrochloride, ethylene carbonate, thiocyanate, dimethyl sulfoxide and urea.
In the process of this invention, the chaotropic auxiliary employed is preferably guanidine, guanidine hydrochloride or particularly preferably, urea.
The concentration of the chaotropic auxiliary in step (b) of the process of this invention is preferably 7.0 to 9M, the temperature in step (b) is preferably 40xc2x0 C. and the pH in step (b) is preferably 10 to 11.
In the process of this invention, the pH in step (c) is preferably 10 to 11. In step (c) of the process of this invention, the amount of water into which the mixture is introduced is preferably selected such that it results in a dilution of the cysteine or cysteine hydrochloride concentration in the mixture to approximately 2.5 to 3 mM and a concentration of the chaotropic auxiliary of approximately 0.5 M.
Particularly preferably, the process according to the invention is one wherein the concentration of the chaotropic auxiliary in step (b) is approximately 8 M, the temperature in step (b) is approximately 40xc2x0 C., the pH in step (b) is approximately 10.6, the pH in step (c) is approximately 10.6 and in step (c) the amount of water results in a dilution of the concentration of the cysteine or of the cysteine hydrochloride in the mixture to approximately 2.5 to 3 mM and a concentration of the chaotropic auxiliary of approximately 0.5 M.
The result of the process of this invention is a precursor of insulin or an insulin derivative, in particular a proinsulin, whose cystine bridges are correctly bonded.
Insulin derivatives are derivatives of naturally occurring insulins, namely human insulin (see SEQ ID NO:1=A chain of human insulin; see SEQ ID NO:2=B chain of human insulin, sequence listing) or animal insulins, which differ by substitution of at least one naturally occurring amino acid residue and/or addition of at least one amino acid residue and/or organic residue of the corresponding, otherwise identical, naturally occurring insulin.
From the precursor of the insulin or insulin derivative having correctly bonded cystine bridges obtained with the aid of this invention, it is finally possible in combination with the process described in EP 0 600 372 A1 (or U.S. Pat. No. 5,473,049) or in EP 0 668 292 A2, to prepare an insulin or an insulin derivative having correctly bonded cystine bridges by enzymatic cleavage by means of trypsin or a tripsin-like enzyme and, if appropriate, additionally by means of carboxypeptidase B and subsequent purification on an adsorber resin.
The insulin or insulin derivative which can be prepared from the precursor can preferably be described by formula I 
in which
Y is a genetically encodable amino acid residue;
z is
a) an amino acid residue from the group consisting of His, Arg and Lys,
b) a peptide having 2 or 3 amino acid residues, comprising the amino acid residue Arg or Lys at the carboxyl end of the peptide,
c) a peptide having 2-35 genetically encodable amino acids, comprising 1 to 5 histidine residues, or
d) OH;
R1 is a phenylalanine residue (Phe) or a covalent bond,
R3 is a genetically encodable amino acid residue,
and where the radicals A2-A20 of the amino acid sequence of the A chain of human insulin (not shown for the simplification of the formula I) correspond to animal insulin or an insulin derivative and the radicals B2-B29 of the amino acid sequence of the B chain of human insulin (not shown for the simplification of the formula I) correspond to animal insulin or an insulin derivative.
The amino acid sequence of peptides and proteins is indicated from the N-terminal end of the amino acid chain onward. The details in formula I in brackets, e.g. A6, A20, B1, B7 or B19, correspond to the position of amino acid residues in the A or B chains of the insulin.
The term xe2x80x9cgenetically encodable amino acid residuexe2x80x9d represents the amino acids Gly, Ala, Ser, Thr, Val, Leu, Ile, Asp, Asn, Glu, Gln, Cys, Met, Arg, Lys, His, Tyr, Phe, Trp, Pro and selenocysteine.
The terms xe2x80x9cresidues A2-A20xe2x80x9d and xe2x80x9cresidues B2-B29xe2x80x9d of xe2x80x9canimal insulinxe2x80x9d are understood as meaning, for example, the amino acid sequences of insulin from cattle, pigs or chickens. The terms xe2x80x9cresidues A2-A20xe2x80x9d and xe2x80x9cB2-B29xe2x80x9d of insulin derivatives represent the corresponding amino acid sequences of human insulin which are formed by the replacement of amino acids by other genetically encodable amino acids.
The A chain of human insulin has the following sequence (SEQ ID NO:1):
Gly Ile Val Glu Gin Cys Cys Thr Ser Ile Cys Ser Leu Tyr Gln Leu Glu Asn Tyr Cys Asn.
The B chain of human insulin has the following sequence (SEQ ID NO:2):
Phe Val Asn Gln His Leu Cys Gly Ser His Leu Val Glu Ala Leu Tyr Leu Val Cys Gly Glu Arg Gly Phe Phe Tyr Thr Pro Lys Thr.
In human insulin, R3 in formula I is asparagine (Asn), R1 is phenylalanine (Phe), Y is threonine (Thr) and Z is OH.
The process according to the present invention is particularly suitable for obtaining a precursor of insulin or an insulin derivative having the formula II, whose cystine bridges (not shown in formula II) are correctly folded,
R2xe2x80x94R1xe2x80x94(B2-B29)xe2x80x94Yxe2x80x94X-Gly-(A2-A20)xe2x80x94R3xe2x80x83xe2x80x83(II),
in which
R2 is
a) a hydrogen atom,
b) an amino acid residue from the group consisting of lysine (Lys) and arginine (Arg), or
c) a peptide having 2 to 45 amino acid residues, comprising the amino acid residue lysine (Lys) or arginine (Arg) at the carboxyl end of the peptide;
R1 is a phenylalanine residue (Phe) or a covalent bond;
(B2-B29) are the amino acid residues in the positions B2 to B29 of the B chain of human insulin, animal insulin or an insulin derivative which is optionally varied in one or more of these positions;
Y is a genetically encodable amino acid residue;
X is
a) an amino acid residue from the group consisting of lysine (Lys) and arginine (Arg),
b) a peptide having 2 to 35 amino acid residues, comprising the amino acid residue lysine (Lys) or arginine (Arg) at the N-terminal end and at the carboxyl end of the peptide, or
c) a peptide having 2 to 35 genetically encodable amino acids, comprising 1 to 5 histidine residues;
(A2-A20) are the amino acid residues in the positions A2 to A20 of the A chain of human insulin, animal insulin or an insulin derivative which is optionally varied in one or more of these positions; and
R3 is a genetically encodable amino acid residue.
1. In a first preferred embodiment, in formula II:
R2 is
a) a hydrogen atom, or
b) a peptide having 2 to 25 amino acid residues, comprising the amino acid residue arginine (Arg) at the carboxyl end of the peptide;
R1 is a phenylalanine residue (Phe);
(B2-B29) are the amino acid residues in the positions B2 to B29 of the B chain of human insulin;
Y is an amino acid residue from the group consisting of alanine (Ala), threonine (Thr) and serine (Ser);
X is the amino acid residue arginine (Arg) or a peptide having the amino acid sequence of the C chain of human insulin;
(A2-A20) are the amino acid residues in the positions A2 to A20 of the A chain of human insulin; and
R3 is an amino acid residue from the group consisting of asparagine (Asn), serine (Ser) and glycine (Gly)
The C chain of human insulin has the following sequence (SEQ ID NO:3):
2. In a second preferred embodiment, in formula II:
R2 is
a) a hydrogen atom, or
b) a peptide having 2 to 15 amino acid residues, at whose carboxyl end is found an arginine residue (Arg);
R1 is a phenylalanine residue (Phe);
(B2-B29) are the amino acid residues in the positions B2 to B29 of the B chain of human insulin;
Y is a threonine residue (Thr);
X is the amino acid residue arginine (Arg) or a peptide having 2 to 35 amino acid residues, where at the beginning and at the end of the peptide there are two basic amino acid residues, in particular arginine (Arg) and/or lysine (Lys);
(A2-A20) are the amino acid residues in the positions A2 to A20 of the A chain of human insulin; and
R3 is the amino acid residue asparagine (Asn) or glycine(Gly).
The residue Z of the insulin or of the insulin derivative of the formula I is, as a rule, part of the amino acid sequence of X of the precursor of the formula II and results from the activity of the proteases such as trypsin, trypsin-like enzyme or carboxypeptidase B, therefor The radical R3 is the amino acid residue which is in position A21 of the A chain of insulin. The radical Y is the amino acid residue which is in position B30 of the B chain of insulin.
Trypsin or trypsin-like enzymes are proteases which cleave amino acid chains at the arginine or lysine residue. Carboxypeptidase B is an xoprotease which removes basic amino acid residues such as Arg or Lys which are at the carboxy-terminal end of amino acid chains. (Kemmler et al., J. Biol. Chem. 246, pages 6786-6791).
From the precursor described above as the first preferred embodiment, it is possible, for example, to obtain an insulin or insulin derivative of the formula I having correctly linked cystine bridges, where Y, R1, R2, R3, A2-A20 and B2-B29 have the precursor described above as a first preferred embodiment, and Z is an argine residue (Arg), a peptide residue Arg-Arg or xe2x80x94OH.
From the precursor precursor described above as the second preferred embodiment, it is possible, for example, to obtain an insulin or insulin derivative of the formula I having correctly linked cystine bridges, where Y, R1, R2, R3, A2-A20 and B2-B29 have the meaning precursor described above as a second preferred embodiment,and Z is an arginine residue (Arg), a peptide residue Arg-Arg or Lys-Lys or xe2x80x94OH.
The precursor of the formula II can be formed in microorganisms with the aid of a large number of genetic constructs (EP 0 489 780, EP 0 347 781, EP 0 453 969). The genetic constructs are expressed in microorganisms such as Escherichia coli or Streptomycetes during fermentation. The proteins formed are deposited in the interior of the microorganisms (EP 0 489 780) or secreted into the fermentation solution.
For the process according to the invention, precursors of insulins or of insulin derivatives of the formula II can be employed that are still contaminated with a large number of proteins which originate from the fermentation solution and from the microorganisms directly after the cell disruption. The precursors of the formula II, however, can also be employed in prepurified form, for example after precipitation or chromatographic purification.